Stanford Prison Experiment

Craig Haney & Philip G Zimbardo. Encyclopedia of Group Processes & Intergroup Relations. Editor: John M Levine & Michael A Hogg. Sage Publications, 2010.

The Stanford prison experiment (SPE) was an experiment designed to examine the power of an institutional environment prison, in particular to shape and control the behavior of persons placed inside it. Using college student participants who were selected for their normality and randomly assigned to be prisoners or guards, the study ended unexpectedly early because of the dramatic and extreme results. It has assumed a prominent place in debates over the causes of extreme behavior in powerful situations or settings, especially in the criminal justice system.

Study Design and Findings

The SPE was conducted in 1971 by a group of Stanford research psychologists, led by Philip Zimbardo, and his two graduate students, Craig Haney and Curtis Banks. The experiment was designed to control for the individual personality variables (e.g., narcissism, authoritarianism) that are sometimes used to attempt to explain behavior in prison and other institutional settings. That is, the researchers in the SPE neutralized the explanatory argument that pathological traits alone accounted for extreme and abusive behavior in severe institutional setting such as prisons. They did this by (a) selecting a group of participants who were psychologically healthy and had scored in the normal range of numerous personality variables and (b) assigning participants to the role of either prisoner or guard on a completely random basis. The behavior that resulted when these otherwise healthy, normal participants were placed in the extreme environment of a simulated prison would therefore have to be explained largely if not entirely on the basis of the characteristics of the social setting in which they had been placed.

The setting itself was designed to be as similar as possible to an actual prison, given a number of obvious practical and ethical constraints. Constructed in the basement of the Psychology Department at Stanford University, the “Stanford County Prison” had barred doors on the small rooms that served as cells, cots on which the prisoners slept, a hallway area that was converted to a prison “yard” where group activities were conducted, and a small closet that served as a short-term “solitary confinement” cell for disciplining unruly prisoners. The prisoners wore uniforms that were designed to deemphasize their individuality and underscore their powerlessness. In contrast, guards donned military-like garb, complete with reflecting sun glasses and nightsticks. Guards generated a set of rules and regulations that in many ways resembled those in operation in actual prisons, and prisoners were expected to comply with the guards’ orders. However, guards were instructed not to resort to physical force to gain prisoner compliance.

Despite the lack of any legal mandate for the “incarceration” of the prisoners, and despite the fact that both groups were told that they had been randomly assigned to their roles (so that, for example, guards knew that prisoners had done nothing to “deserve” their degraded prisoner status, and similarly, prisoners knew that the guards had no special training or actual legal authority over them), the behavior that ensued was remarkably similar to behavior that takes place inside actual prisons. It also was surprisingly extreme in intensity and effect. Thus, initial prisoner resistance and rebellion were met forcibly by guards, who quickly struggled to regain their power and then proceeded to escalate their mistreatment of prisoners throughout the study, at the slightest sign of affront or disobedience.

As the guards’ control over the prisoners increased, the tensions between the two groups intensified, and the harassment of the prisoners worsened. For example, the guards conducted a series of “counts” times when prisoners were removed from their cells in order to be counted but which quickly deteriorated into occasions for verbal and other forms of abuse and humiliation that the guards directed at them. In some instances, the guards conspired to physically mistreat prisoners outside the presence of the experimenters and to leave prisoners in the solitary confinement cell beyond the 1-hour limit that the researchers had set.

Conversely, prisoners resisted the guards’ orders at first but then succumbed to their superior power and control. Some prisoners had serious emotional breakdowns in the course of the study and had to be released from participation. Other prisoners became compliant and conforming, rarely if ever challenging the “authority” of the guards. Despite the fact that the researchers could not keep the prisoners in the study against their will (and they had been informed at the outset of the study of their legal right to leave), as the study proceeded prisoners “petitioned” the prison “administrators” for permission to be “paroled” and returned passively to their cells when their requests were denied. By the end of the study, they had disintegrated as a group. The guards, in contrast, solidified and intensified their control as time passed. Although some of the guards were more extreme and inventive in the degradation they inflicted on the prisoners and some were more passive and less involved, none of the guards intervened to restrain the behavior of their more abusive colleagues. Although the study was designed to last for 2 full weeks, the extreme nature of the behavior that occurred led the researchers to terminate it after only 6 days.

A post hoc analysis of the SPE data showed that the careful screening of the participants and their random assignment to the roles of prisoners and guards had effectively controlled for any significant personalitybased, or dispositional, explanation of the results. That is, there were no significant personality differences between the SPE participants and the normal population (i.e., the group means for guards and prisoners did not fall outside the 40 to 60 percentile range of the normative male population on any of the dimensions of the personality inventory that was used), no personality differences between the guards and prisoners could explain their very different behavior in the study, and no personality differences within either group reliably predicted variations in their in-prison behavior.

The SPE’s Larger Implications

Controversial from the outset and widely discussed since it was conducted, the study has come to stand in psychology and related disciplines as a demonstration of the power of situations especially extreme institutional settings such as prisons to shape and control the behavior of the persons in them. The results of the study undermine the notion that extreme social behavior can only or even mostlybe explained in terms of the extreme characteristics of the persons who engage in it. Instead, the SPE warns us to look more carefully at the characteristics of the settings in which extreme behavior occurs.

The SPE also stands as a challenge to what might be termed the presumption of institutional rationality that is, the tendency to assume that institutions operate on the basis of an inherent rationality that should be accepted rather than questioned. Instead, the SPE (itself the most “irrational” of prisons, in the sense that the guards had no legal authority over the prisoners, who had committed no crimes that warranted their punishment) suggests that a kind of “psychologic” may operate in these settings that controls role-bound behavior, whether or not that behavior furthers legitimate goals. That is, despite the fact that the guards had no genuine authority over the prisoners and the prisoners had done nothing illegal to legitimize their mistreatment, the guards reacted to violations of rules that they arbitrarily constructed as if they were mandated to do so (and often did so forcefully, in ways that caused apparent pain and distress for the prisoners).

The SPE was conducted in the early 1970s, and its results and implications were widely disseminated in the years that immediately followed. The study was often cited as a prominent example of research that contributed directly to the “situational revolution” in psychology the insight that context plays a powerful role in shaping people’s thoughts and actions, especially in extreme settings (such as ones where social pressures are brought acutely to bear, where marked imbalances of power exist, and where all aspects of patient, inmate, or prisoner behavior are subject to control). It helped lead to the now more widely accepted proposition that no account of behavior can avoid a careful assessment of the situational influences on what people do, the things they believe, and even how they think.

The study also promised to have an impact on prison policy, at least in the years that immediately followed. This was a time in the history of the criminal justice system in the United States when the nation appeared open to fundamental reform of its crime control policies and penal practices. The message of the SPE that context very much mattered in general and that, specifically, prison and prisonlike environments had the inherent capacity to set powerful social-psychological forces in motion that could negatively affect the behavior of staff members and have adverse consequences for inmates resonated perfectly with the spirit of these times.

However, for reasons that appeared more political than scientific, the nation’s prison and crime control policies soon began to move in a fundamentally different direction. In the waning years of the 1970s and in the several decades that followed, the situational message of the SPE had little impact in correctional circles. Although evidence continued to mount in psychology and related disciplines that past and present circumstances and situations played a powerful role in influencing behavior, not only in penal institutions but in the origins of the criminal behavior those institutions were intended to address, sentencing laws and prison policies were implemented that seemed to ignore the most important lessons of the situational revolution. That is, crime control practices during these years focused even more narrowly on individual-level wrongdoing to the exclusion of situational models of crime prevention, and the potential role of context and circumstance in crime causation was increasingly discounted, even in sentencing guidelines (where “social factors” were explicitly deemed irrelevant).

In more recent years, however, the implications of the SPE became part of the national dialogue that occurred in response to the widely publicized abuse of prisoners at Abu Ghraib, Iraq, by members of the U.S. military. There appeared to be direct parallels between some of the mistreatment perpetrated by the guards in the SPE and the abuse that was inflicted on the prisoners at Abu Ghraib. Here, too, the explanation that situational forces had overcome the dispositions of the otherwise normal, healthy soldiers who perpetrated the abuse seemed cogent. The analysis of the behavior of the guards in the SPE and at Abu Ghraib pointed away from a “few bad apples” assessment of blame and focused instead on the abuse-engendering circumstances in which the guards functioned a “faulty barrel” assessment, if you will as well as the responsibility of the persons who created and maintained such a flawed environment for monitoring prisoners and guards alike.